Background: Fifty percent of medications are not taken as prescribed. This is a major public health issue yet there is very limited evidence on the factors associated with medicines adherence in people with mild to moderate Intellectually Disabilities and diabetes (IDD). This study evaluated the frequency of, and factors associated with, medicines non-adherence in this group compared to people without ID but with diabetes (non-IDD).Methods: A systematic review of the literature informed the theoretical model tested. A two-stage, sequential mixed methods study with 111 people with type 1 and 2 diabetes, (IDD = 33, non-IDD = 78) was then carried out. Stage one (quantitative) compared frequency of medication adherence in the group overall, IDD and non-IDD. Univariate and multiple regression analysis evaluated associations between factors (ID, depression, side effects, self-efficacy and perceived level of social support) and medicines non-adherence. Stage two (qualitative) explored findings of stage one with 12 stage one participants' carers using semi-structured interviews. Results: Data were collected between July 2014 and May 2016. The frequency of medicines adherence was similar in the IDD and non-IDD population (70% vs 62%, p = 0.41). The theoretical model did not predict medicines non-adherence. After controlling for support with medicines and complexity of regime (number of medications and use of insulin), depression was an independent predictor in the non-IDD and group overall (p < 0.001). In the IDD group, perceived side effects was an important, but non-significant, predictor of non-adherence (p = 0.06). Carers' perceptions of adherence and depression were consistent with stage one findings. Conclusions: Optimising adherence to diabetes medicines is equally challenging in IDD and non-IDD populations. Associations between independent factors and adherence differed between the two groups: in the non-IDD population, depressive symptoms were associated with non-adherence whereas in the IDD population perceived level of side effects appeared most dominant. Due to small sample sizes findings were inconclusive therefore, a sufficiently powered study further investigating the relationship between adherence and side effects in people with ID and diabetes is recommended.